Genetically flexible murine organoids for mechanistic and functional modeling of cholangiocarcinoma

Cholangiocarcinoma (CCA) is the second most common liver cancer and its incidence continues to rise. The 5-year survival rate remains below 10%, mainly because of late diagnosis and the lack of efficient therapeutic strategies. Several studies have identified genetic alterations in human CCA, but their effect on development, drug resistance and tumor progression is mostly unclear. Therefore an appropriate murine model system is required that resembles its human counterpart. Human CCA frequently exhibits CK19-positive cancer cells growing in ductular structures surrounded by a desmoplastic stroma [1]. Liver organoid cultures were recently established from human CCA, Hepatocellular Carcinoma (HCC) and normal liver tissue [2]. Advantages of the organoid culture system are the commitment to the tissue of origin and a persistent genetic profile even after long term culture.

We developed and characterized a genetically tractable murine CCA model, which can be used for studying the multi-step process of tumorigenesis in immunocompetent mice. Murine liver organoids were established and genetically modified using the Cre-lox system, RNAinterference and CRISPR-Cas9 technology. Liver organoids carrying an endogenous oncogenic K-Ras allele in combination with loss of the tumor suppressor p53 lead to tumor formation in recipient mice with 100% penetrance. These tumors were classified as moderately differentiated adenocarcinomas with CK19 positive cancer cells surrounded by a stromal reaction. Loss of the tumor suppressor Pten showed acceleration of tumor growth in-vivo. Treatment with Gemcitabine led to only a moderate survival benefit in mice suggesting that the model can be used to study chemotherapy response. To make the system scalable, 2D and 3D (tumoroids) cancer cell lines from organoid derived tumors were established. These cultures can be serially transplanted and show similar histological features as the primary malignancy and will be useful tools for further higher throughput approaches.

Liver organoids exhibit a biliary marker phenotype in vitro. Still, in-vivo modified organoids can either induce tumors with a HCC-like phenotype or CCA, illustrating the plasticity of organoids depending on the genetic background. These phenotypes were further characterized via transcriptome analysis. Both cancer-types show a genetic profile comparable with previously published respective human data sets.

Conclusion:

We show that genetically modified murine liver organoids serve as a histologically accurate and genetically flexible in-vivo system for liver cancer development. Furthermore, it can be used to characterize cancer drivers and analyze the effect of genetic changes on the stromal reaction. The resulting tumors can be induced in immunocompetent mice with a defined genetic background and show a transcriptomic profile comparable to human HCC or CCA, respectively, a perfect prerequisite to study novel therapeutic approaches.

[1] J. Banales, V. et al: Cholangiocarcinoma: current knowledge and future perspectives consensus statement from the European NEtwork for the Study of Cholangiocarcinoma (ENS-CCA),“ Gastroenterology & Hepatology, May 2016.

[2] L. Broutier et al: "Human primary liver cancer-derived organoid cultures for disease modeling and drug screening.," Nature Medicine, pp. 1424 – 1435, 23 December 2017.